There have been several methods available to estimate speed of a vehicle while in motion. Generally, the speed of the vehicle is measured using an On-Board diagnostics system (OBD). For example, the OBD system may have a plurality of sensors such as an accelerometer, and a GPS to measure the speed. In recent times, portable electronic devices are used to measure the speed. Specifically, the portable electronic devices comprise the plurality of sensors to measure the speed. When using the portable electronic devices, use of the sensors such as the GPS sensor may drain battery of the portable electronic devices. Further, the GPS sensors may not estimate the speed accurately when the portable electronic devices are moving in a tunnel Further, when the portable electronic devices are carried in the vehicle when moving between various geographies, the GPS sensor may not estimate the speed accurately due to poor reception of signals from a satellite.
In addition, the GPS sensors estimate the speed at low sampling rates. In one scenario, the GPS sensor may measure the speed at a time interval of 10 seconds. In another scenario, the GPS sensor may measure the speed at the time interval of 5 seconds. In order to estimate the speed accurately in a granular level, several methods have been proposed. One such approach includes fusing values of the GPS sensor with values received from an inertial sensor. However, when the vehicle is moving, the values measured using the above approach may have several errors such as bias, noise etc. The errors may degrade estimation of the speed resulting in large divergence in speed estimation.